Hot Off the Press – March 18, 2026
Published in PNAS by Ying Duan, Pei-Jung Tsai and Yihong Yang, et al. of the NIDA IRP Magnetic Resonance Imaging and Spectroscopy Section.
Summary
This study is a trans-branch collaboration between groups of Drs. Yihong Yang and Yavin Shaham at NIDA-IRP. Relapse driven by craving remains a challenge in treating opioid use disorder (OUD). In this study, we used longitudinal fMRI and machine-learning based predictive modeling to identify a whole-brain connectivity pattern that predicts incubation of oxycodone craving in rats after voluntary abstinence. Connectivity changes during abstinence successfully predicted craving magnitude, measured by changes in lever pressing under extinction testing. The predictive connectivity pattern encompassed circuits across cortical, basal ganglia, insular, hippocampus and sensorimotor systems related to reward, motivation, stress and habit formation. Pharmacological inactivation of the dorsal medial striatum, a hub in the connectivity network, altered the connectivity pattern, demonstrating its causal and treatment-sensitive nature. These findings reveal a brain-wide marker of craving and provide a translational framework for developing potential interventions targets and monitoring therapeutic outcomes in OUD.
Publication Information
Predicting individual incubation of opioid craving by whole-brain functional connectivity Journal Article
In: Proc Natl Acad Sci U S A, vol. 123, no. 11, pp. e2531921123, 2026, ISSN: 1091-6490.

